Recently, a large number of measurements using intermolecular interactions such as immune responses are being carried out in clinical tests, etc. However, since conventional methods require complicated operations or labeling substances, several techniques are used that are capable of detecting the change in the binding amount of a test substance with high sensitivity without using such labeling substances. Examples of such a technique may include a surface plasmon resonance (SPR) measurement technique, a quartz crystal microbalance (QCM) measurement technique, and a measurement technique of using functional surfaces ranging from gold colloid particles to ultra-fine particles. The SPR measurement technique is a method of measuring changes in the refractive index near an organic functional film attached to the metal film of a chip by measuring a peak shift in the wavelength of reflected light, or changes in amounts of reflected light in a certain wavelength, so as to detect adsorption and desorption occurring near the surface. The QCM measurement technique is a technique of detecting adsorbed or desorbed mass at the ng level, using a change in frequency of a crystal due to adsorption or desorption of a substance on gold electrodes of a quartz crystal (device). In addition, the ultra-fine particle surface (nm level) of gold is functionalized, and physiologically active substances are immobilized thereon. Thus, a reaction to recognize specificity among physiologically active substances is carried out, thereby detecting a substance associated with a living organism from sedimentation of gold fine particles or sequences. Surface plasmon resonance (SPR), which is most commonly used in this technical field, will be described below as an example.
A commonly used measurement chip comprises a transparent substrate (e.g., glass), an evaporated metal film, and a thin film having thereon a functional group capable of immobilizing a physiologically active substance. The measurement chip immobilizes the physiologically active substance on the metal surface via the functional group. A specific binding reaction between the physiological active substance and a test substance is measured, so as to analyze an interaction between biomolecules. An example of a surface plasmon resonance measurement device is the device described in Japanese Patent Laid-Open (Kokai) No. 2001-330560.
When a specific binding reaction between a physiologically active substance and a test substance is measured, the binding reaction is generally measured by: connecting in series a reference cell, to which a physiologically active substance interacting with a test substance does not bind, with a detection cell, to which a physiologically active substance interacting with a test substance binds; placing the connected cells in a flow channel system; and feeding a liquid through the reference cell and the detection cell, so as to carry out the measurement of the binding reaction. During the measurement, the liquid contained in the above flow channel system is exchanged from a reference liquid containing no test substance to be measured to a sample liquid containing a test substance to be measured, so as to cause the binding reaction between the physiologically active substance and the test substance to be initiated, and to measure a change in signals due to a lapse of time.
As mentioned above, a biosensor that uses surface plasmon resonance detects the binding of an analyte to a sensor (a metal film and a ligand) as a change in the refractive index (and an angular change of a dark line caused thereupon). If the time is plotted on the horizontal axis and the binding signal is plotted on the vertical axis, a signal (indicating the amount of binding or the like) that is referred to as a so-called “sensorgram” can be observed with the elapse of time (see, for example, FIG. 3). It is important to carry out fitting of the following rate equation (i) to the sensorgram followed by determination of the rate coefficients such as the adsorption rate coefficient (Ka) and the dissociation rate coefficient (Kd). Such procedures are extensively adopted in the field of drug screening.dR/dt=Ka×C×{Rmax−R(t)}−Kd×R(t)  (i)R(t)=(Ka×C×Rmax)/(Ka×C+Kd)×(1−exp(−Ka×C+Kd)×t))  (ii)(the result of solving equation (i))wherein Ka represents an adsorption rate coefficient; Kd represents a dissociation rate coefficient; C represents an analyte concentration (known); Rmax represents the theoretical maximum amount of binding; and t represents a time.
As is apparent from equations (i) and (ii), the Rmax value was necessary in order to determine Ka and Kd based on the sensorgram showing the binding.
Thus, the following measurement or other measurement techniques had been carried out in the past:
(a) performing a measurement wherein analyte concentrations are set at different levels (FIG. 4);
(b) measuring the amount of ligands immobilized in advance and calculating the Rmax value based on the measured value (FIG. 5); and
(c) performing nonlinear regression analysis by employing Rmax as a variable as with Ka and Kd. However, it is cost-consuming and time-consuming to carry out the measurement (a) due to an increased number of measurement operations. The measurement (b) was also problematic in terms of a prolonged measurement time, an increased amount of labor, and deterioration in measurement accuracy due to errors resulting from variations in the amount of ligands immobilized. When nonlinear regression analysis is performed as in the case of the (c) above, detection accuracy was also deteriorated due to errors resulting from increased computational effort and errors resulting from variations in the amount of ligands immobilized.